Comparative Morphology of Two Sympatric Pareledone Species from South Georgia

Antarctic Science 6 (2): 163-169 (1994) Comparative morphology of two sympatric Pareledone species from South Georgia

HEATHER I. DALY 1~2and PAUL G. RODHOUSE1 'BritishAntarctic Survey, Natural Environment Research Council, High Cross, Madingley Road Cambridge CB3 OET, Uh: 2Department of Zoology, University of Aberdeen, Tillydrone Avenue, Aberdeen AB9 2TN, UK

Abstract: Morphometricdata were collected for 410 specimens ofpareledone turqueti and P. polymorpha caught around South Georgia. The two species differ in beak morphology and in the male hectocotylus. The species have similar appearances although there is a small but significant difference in the mantle lengthbody mass relationship for females, with P. polymorpha having a relatively longer mantle. There is no significant difference in the arm lengthbody mass relationship between species or sexes (p > 0.05), except in the case of arm IV of females. There is an interspecific significant difference between sucker number on arms I and I1 of males, arms I-IV of females, and between hood length and mass of the buccal mass (p c 0.05), with P. turqueti having relatively lower sucker numbers, a longer hood length and greater buccal mass mass. The beak of P. turqueti is similar to that of Eledone spp. but P. polymorpha has a small, fine beak with the rostra1 tip ending in an elongated, sharp point. Differences in beak and buccal mass suggest that these sympatric species occupy distinct trophic niches and that the differing morphology of the male hectocotylus is a factor in reproductive isolation.

Received 5 July 1993, accepted 3 September 1993 Key words: Pareledone spp., South Georgia, sympatry, beaks, hectocotylus

Introduction Materials and methods On the South Georgia shelf, two closely related species of Between 6-29 January 1987,seven Aggasiz trawls by RFLS John octopodids, Pareledone turqueti (Joubin 1905) and BiscoeinCumberlandEastBay,South Georgiapig. 1)collected P. polymorpha (Robson 1930),occur together. The two species 67 specimens of Pareledone which were preserved and stored are distinguishable by their beak morphology and by the in c. 5% formalin on capture. Voight (1991) validated the use hectocotylised third right arm tip of the males. Few data exist of preserved specimens for morphology studies using a number on the ecology of Pareledone, as literature is restricted to of octopus collections. Bottom depthsfor the trawls rangedfrom descriptionsof preserved specimensand records of geographical 131-265 m and the haul duration varied from 10-35 minutes. distribution (Massy 1916, Berry 1917, Robson 1932), or to The sampleswere identifiedas thespeciesP. turqueti,comprising accounts of gonad size (Kuehll988). Dell (1972) recognized 29 males and 19 females, and P. polymorpha comprising 11 P. turqueti as circum-Antarctic, and there is one recorded males and eight females. Measurements of total body mass specimen from Rio de Janeiro (Dell 1959), the only reported (BM), dorsal mantle length (ML), arm length (AL) and sucker instance of this species north of the Antarctic Polar Front. The number (SN) for the four arms on the right side, beak hood full distribution of P. polymorpha is not known. Kuehl(l988) length (HL), and buccal mass mass (BMM) were taken for all recorded P. polymorpha as slightly more abundant than specimens, and ligula and calamus lengths recorded for males. P. turqueti during a trawl survey on the shelf around the South ML, AL, HL, and ligula and calamus lengths were recorded in Shetland Islands. The occurrence of these two closely related mm, and BM and BMM in g. These standard parameters for benthic species in the same area suggests that they occupy describing cephalopod anatomy were selected from Raper & discrete trophic niches and have divergent reproductive cycles. Voss (1983). This hypothesis is supported by the markedly different beak Thebeaksof eachspeciesweredescribed indetailafter Clarke morphology and relative buccal mass size of the two species, (1986). Cephalopod beaks are useful in dietary studies due to along with comparatively higher numbers of arm suckers in their resistance to digestion, and tendency to collect in the P. polymorpha. This study quantifies some morphological stomach of predators, and the lower beak is usually described as similaritiesanddifferencesbetweenthesetwosympatricspecies, it has more distinguishing features. The funnel organ of a and examinesthe possibility that theirspecialised anatomiesare subsample was also examined for W or VV shapes (Robson related to different trophicniches,enabling them to avoid direct 1932). competition through ecological separation. Trawls by the MV Falklands Protector on the shelf around South Georgia (Fig. 1)between 4-9 February 1989 collected a sample ofpareledone spp. for size frequency analysis (n-343).

163

Downloaded from https://www.cambridge.org/core. Open University Library, on 08 May 2019 at 13:35:28, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0954102094000258 164 H.I. DALY and P.G. RODHOUSE

Fig, 1. Map of South Georgia, showing stations where Pareledone turqueti and P. polymorpha were caught. (Half-filled circles indicate stations where both species were caught, filled circles where only P. turqueti was caught and open circles where only P, polymorpha was caught).

The samplecomprised 119 male and 108femaleP. turqueti, 52 package (Ryan et al. 1985), and all data presented in graphical male and 64 female P,polymorpha, and one female P. charcoti form were logc transformed. The preservation process of the (Joubin 1905),andwas frozenimmediatelyon capturefor return sample for morphological analysis meant that some mantle to the UK. Sex, species, total body mass and dorsal mantle shapes were slightly distorted, so body mass, rather than mantle length were recorded from this sample. Additional species length, was used as the independent variable throughout to differences in skin colour, body consistency and muscularity derive allometric relationships. were noted. Statisticalanalyses were performed using theMinitab software Results Cumberland East Bay and the position of hauls around South I Pareledone furqueti d Pareledone turqueb Q Georgia are shown in Fig. 1. In the combined 1987 and 1989 tI collections 67% of the catch was composed of P. turqueti, and 33% ofP.polymorpha, with a singleP. charcoti specimen. The size range of P. turqueti was 22-107 mm ML and &211 g BM and the size range of P. polymorpha 16-113 mm ML and 4-274 g BM. The P. charcoti was 50 mm ML and 133 g BM. The size frequency distribution of the 1989 collection is shown in Fig. 2. T-tests assessed significance of difference between distributions.There is no significantdifferencebetween the mean ML of male (63.6 k 11.5) and female (66.8 k 15.2) P. turqueti (t= 1.75,p>0.05). ThemeanMLofP.polymorpha females (78.3+ 19.3) is significantly larger (t = 4.17,~< 0.001) than the males (65.9 k 12.5). Comparison between the males of both species shows no significant difference (t = 1.14, p > 0.05). The P. polymorpha females have a significantly larger mean ML than the P. turqueti females (t = 4.09, I p < 0.001). Pareledone Poreledone polyrnorpha d polymorpho 0 Comparative external morphology.

-- Fig, 2. Size frequency distribution of Pareledone turqueti and P. P. polymorpha having a pink tinged skin and a pale blue polynwrpha, (n = 343). broken border around the lateral edge between the dorsal and

Mantle length vs Body Mass

"1 d

I0 20 30 40 50 log. Body Mars

0 hmledone turquetz 0 Ponkdonepolymorpho

Fig. 4. Relationship between mantle length and body mass for Pareledone turqueti and P. polymorpha.

Right Arm lengths vs Body Mass

I I 9

Fig. 3. The hectocotylus tip of Pareledone turqueti and P.polymorpha. a = P. turqueti, b = P.polymorpha. ventral mantle surfaces. P. rurqueri is greybrown, and lacks

the defined border around the mantle edge. P. turqueti 0 specimens were muscular, and resistant to abrasion, as 3 10 20 30 40 50 10 20 30 40 50 specimens were rarely damaged. P. polymorpha is delicate log. Body Moss and less muscular and many specimens were badly damaged 0 Fbbreledone turqucli o Fbnkdonepdymorpho by abrasion during trawls, with extensive loss of skin and suckers. The funnel organs of the preserved specimens Fig. 5. Relationship between arm length and body mass for showed some morphology but it was not clearly present in all Pareledone turqueti and P. polymorpha. specimens. Of the 17 P. polymorpha examined, 12 could be interpreted as W shapes, two had unrecognizable patterns, The tip of the hectocotylised armofmales is morphologically and three had no patterns. Of 36 P. turqueti funnels distinct for each species (Fig. 3). P. turqueti has a smooth examined, all had varying patterns of ridges, four of which ligula depression and P. polymorpha has a ligula with could be described as VV shaped, with the remaining 32 transverseridges. 1nP.poZymorpha theligulais 13.1%of the having no definite form. W or W shapes were not apparent length of right arm 111 and the calamus 51.1%of the ligula. in the funnel organ of frozeddefrosted specimens. For P. turqueti the mean values are 7.9% of the arm length

Sucker Number vs Arm Length Table I. Summary of results on analysis of covariance (ANCOVA) between I d I P body mass (BM) and mantle length (ML), arm length (AL), hood length (HL) and buccal mass mass (BMM), and AL and sucker number (SN) in Pareledone turqueti and Pareledone polymorpha. (Significance level p < 0.05, NS = not significant).

Comparison Slope Intercept 730 0 P< (F> P< Males ML v BM 0.50 NS -0.63 NS ALIvBM -0.97 NS -0.70 NS ALIIvBM -0.95 NS -0.50 NS AL I11 v BM 0.02 NS -0.90 NS ALIVvBM -1.00 NS 3.55 NS 30 SNvALI 5.83 0.025 40.38 0.001* SNvALII 12.12 0.01 56.00 0.001* SNVAL I11 -0.22 NS 0.25 NS SNvALIV 2.25 NS 3.00 NS Females MLvBM 0.34 NS 5.36 0.05 ALIvBM -0.94 NS 0.41 NS ALIIvBM 0.28 NS -0.82 NS AL 111 v BM -0.95 NS -0.58 NS ALIVvBM 4.72 0.05 -1.05 NS* SNvALI 6.67 0.025 21.67 0.001' SNvALII 0.17 NS 6.25 0.025 SNVAL I11 0.28 NS 11.23 0.01 SNvALN 3.40 NS 23.63 0.001 Combined sexes HL v BM 5.83 0.025 321.17 0.001* BMM v BM 3.56 NS 235.62 0.001 30 35 40 IS 50 30 35 do 45 50 --log. Arm Length *Where a significant difference is seen between the slopes the comparison, Poreledone tvrqvefr o breledonepol~morpho of the analysis for a significant difference in intercept assumes a common slope. Fig. 6. Relationship between sucker number and arm length for Pareledone turqueti and P. polymorpha. for females were not significantly different in slopes but the differencebetween interceptswas significant(TableI),indicating and the calamus 36.0% of the ligula. a higher number of suckers for a given arm length in The ML, AL, HL and BMM plotted against BM for both P. polymorpha. species are shown in Figs 4,5,8 respectively. Fig. 6 shows the relationship between SN and Analysis of covariance AL. Feeding structures (ANCOVA)wasused to test for interspecificdifferencesin these relationships and results are summarized in Table I. The The lower beaks of both species are illustrated in Fig. 7. In relationship between ML and BM (Fig. 4) shows that the two the following description the term 'normal' refers to the beak species and both sexes appear similar in their overallproportions. shape characteristic of the subfamilies Octopodinae, There was no significantdifference between theslopesfor males Eledoninae and Bathypolypodinae (Clarke 1986). or females, or in theinterceptsforthemales (TableI). However, The P. turqueti beak described and illustrated was from a the intercepts for the MLDM relationship were significantly female specimen with a ML of 97 mm and a BM of 187 g, and different in females, with P.poZymorpha having a slightly larger the P.polymorpha beak from a female with a ML of 80 mm and ML when the effect of BM was removed. The relationship a BM of 134 g. Specimens of both species were at a late stage between AL.and BM (Fig. 5) shows no significant difference in of sexual maturity. the slopes or intercepts for any male arms or arms 1-111 of The general shape and relative length of the hood are the most females (Table I). Arm IV in the females shows a small but prominent differences between the species. The hood of significant difference between the slopes for the species. In the P. turqueti was 5.05 mm long and darkened from the rostra1 tip relationshipbetween SN and =(Fig. 6), there was a significant back 4.27 mm with a narrow whitehranslucent border along the difference in the slopes of arms I and I1 for males and arm I for posterior edge. The chitin is dark over most of the beak, the females (Table I). Sucker numbers on arms 111 and IV for males hood, rostrum and crest beingcolouredblack, and the wings and were not significantly different. Sucker numbers on arms 11-IV lateral walls dark brown. The wings and lateral walls have a

Table 11. Comparison of ratios of beak dimensions of Parelodone turqueti and P. polymorpha, parameters a-f from Clarke (1986), a = rostral edge, b = wing length, c = height, d = base length, e = wing width, f = crest length and g = hood length.

Ratio P. turqueti P. polymorpha wing lengthhostral edge (b/a) 5.30 2.27 height%ase length (dd) 0.83 0.78 base lengthhving width (d/e) 2.41 2.65 crest lengthhood length (Vg) 1.91 2.40 base lengthhest length (d/f) 1.98 0.95

wing or jaw angle of P. turqueti is close to right-angled and in cross-section the crest is narrow. The wing angle of P. polymorpha is obtuse and in cross-section the crest ir; broad. Large P. turqueti beaks often have a small indentation in the rostral tip, probably due to wear, which isnot apparentin :smaller specimens which have rounded rostra. The rostral tip of

2mm P,polymorpha beaks are sharp and pointed, and ,have a characteristic slight upturn, which is more pronounced in smaller individuals. Ratios describing beak dimensions, after Clarke (1986), are shown in Table 11. The ratios aid discrimination between some species. The wing lengthhostral edge @/a) ratio in most Octopodidaeis> 4 (Clarke 1986). InP. turqueti b/a = 5.30, but forP.polymorpha b/a = 2.27. The base lengthkrest length ratio (d/f) is larger in P. turqueti (Table II), indicating that it has a proportionately longer base and shorter crest. Ratios height/ base length (c/d), base length/wing width (d/e) and crest length/ hood length (Ug) are not notably different between the spcies. The relationship between HL and BM for each species is shown in Fig.8a. There is a significant difference between the Fig. The lower beak of Pareledone turqueti and P. polymorpha. 7. slopes and intercepts (TableI) indicating differentrelationships, a = P. turqueti, 97 mm ML, 187 g BM. b = P. polyrnorpha, 80 mm ML, 134 g BM. with a relatively longer HL in P. turqueti. Fig.8b shows the relationship between BMM and BM. There is no significant difference between slopes but there is a significant difference narrow white border around the periphery. The hood of between the intercepts (Table I). P. polymorpha was 3.37 mm long and darkenedfrom the rostral tip back to c. 3.09 mm, gradually becoming transparent at the Discussion posterior edge. The chitin is black around the hood and rostrum and brown chitin extends a short way down the wings, just past The two Pareledone species from the South Georgia shelf are the free biting edge of the beak. The main part of the wings are superficially similar, although the P. polymorpha females are white/translucent. The lateral walls are brown and have a larger than the males (Fig. 2) and have a slightly longe,r ML narrow white border aroundthe edges. TheP. turquetibeakhas relative to P. turqueti females (Table I). The AL to BM a prominent rostrum, a curved rostral edge and the hood is a relationships are very similar, with only the fourth aim of ‘normal’ distance from the crest. TheP.polymorpha beak has females showing any significant difference between species. a less prominent rostrum, a straight rostral edge and the hood Identification ofpareledone species has relied on characteiistics lies close to the crest. The hood-wing structure is ‘normal’ such as the hectocotylus (Dell 1959, Nesis 1987) which is only relative to the lateral wall inP. turqueti, and is fairly narrow in useful in males, and in the funnel organ morphology (Massy P, polymorpha. The rostrum protrudes a ‘normal’ distance 1916, Berry 1917). This study found some patterns in the forward inP. turqueti and a top view of the beakshows amarked funnels of both species, but interpretation of the shapes formed V-indentation between the lateral walls, from the posterior is ambiguous in some cases. No funnel organ shapes were corners to the crest. The P. polymorpha beak has a rostrum apparent in three P. polymorpha funnels, indicating the which protrudes forward a long distance and the topview shows unreliability of the funnel organ as a means of supporting a slight V-indentationbetween the lateral walls, although some identification. TheP. turqueti funnel organs often had narrow specimenshave aslightly moresquare-shaped indentation. The ridges on the inner surface but the patterns formed could not be

females, the relatively higher number of suckers along arms a Hood Length vs Body Mass I-IV in females and arms I and I1 in males of P. polymorpha (Table I) means the suckers may be closer together or have a 1.80 smaller diameter. Numerous small or closely packed arm suckers may be an adaptation to manipulate small prey items. .f In males arm I11 on the right side forms the hectocotylus and is t3) 5 1.20 usually the shortest and has fewest suckers. 2 " The beak of P. turqueti is similar in overall shape to Eledoize U 0 ODo cirrhosa (Lamarck) and to Octopus vulgaris(Cuvier) which are 0 - U illustrated in Clarke (1986), and is a common beak form among I, 0.60 - 0 cn 0 octopods. The beak of P. polymorpha is comparatively small -0 andfine(Fig.7), andissirnilartothebeakofl'. adelieana(Berry 1917) and P. umitakae (Taki 1961). Throughout different o.oot 8 I I I I animalgroups,such specialization of mouthparts among closely 1 .o 2.0 3.0 4.0 5.0 related species indicates trophicniche separation (Giller 1984). log, Body Mass Cephalopod diets are not well known but existing information indicates that they have similar, varied diets, consisting mainly 0 F? turqueti males o F? polymorpha males of benthic molluscs, crustacea and fish with polychaetes, chaetognaths and siphonophorestaken by some species (Bidder w F! turqueti females F? po/ymorpha females 1966, Nixon 1987). Beak shape suggests that the diet of P. turquetihassimilarcomponentstootheroctopus species with b Buccal Mass Mass vs Body Mass a 'normal' beak. The beak of P. polymorpha is more delicate and fine, and is unlike any other octopod beak illustrated in Clarke (1986). This species beak is possibly specialized to exploit an atypical resource, possibly in the water column. The delicacy ofP. polymorpha, indicated by the less muscular body, higher number of small suckers and frequency of damage among trawled specimens, might also indicate adaptation to a life style hunting off the bottom. The robust body ofP. turqueti, on the other hand, might indicate amore conventionalbehaviour for octopuses, hunting wholly benthic prey. The diet of the two Pareledone species, identified from 0 stomach contents, willbethesubject offuture research, although 0 this approach may give biased results. A degree of external 00 digestion takes place in the crustacean prey of 0. vulgaris I I I I 1 .o 2.0 3.0 4.0 5.0 (Nixon 1984) andE. cirrhosa (Grisley & Boyle 1990), making log, Body Mass it difficult to identify soft material in the stomach. Other approaches, such as experimental aquarium studies, are required Fig. 8. a. Relationship between hood length and body mass for to relate the anatomical adaptations, in mouthparts and sucker Pareledone turqueti and P.polymorpha. b. Relationship number, to behavioural and life cycle differences. between buccal mass mass and body mass for P. turqueti and P.polymorpha. Acknowledgements easily defined. We thank E.M.C. Hatfield for her advice in statistical analysis Inter-breeding between these similar species is prevented by and R. Missing for technical assistance. H.I.D. is funded by reproductive mechanisms. Genetic isolation is apparently NERC (CASE) studentship GT4/92/104 between British assured by the differingrelativesizeandshape of the hectocotylus, Antarctic Survey and Aberdeen University Zoology Department which is approximately 50% larger in P. polymorpha than in and is supervised by Prof. Peter Boyle. P. turqueti. There also appeared to be sexual dimorphism in P. polymorpha, with the modal ML of females 20 mm longer References than males (Fig. 2). Sexual dimorphism was not apparent in P. turqueti. The body mass of a female octopus tends to be B~Y,S.S. 1917. Cephalopcda.Austra1ianAnfarctic Expedifion 1911-1914. Scientific Repor?, Series C: Zoology and Botany, 4(2), 1-38. proportional to the number of eggs she lays (Mangold 1987). BIDDER,A.M. 1966. Feeding and dgestion in cephalopods. In WUUR, K.M. & As there was no significant interspecific difference in the YONGE,C.M., eds. Physiology of Mollrcsca 2. London: Academic Press, relationship between AL and BM in all cases except arm IV of 97-124.

CLARKE,M.R. ed. 1986. A handbook for the identification of cephalopod NIXON,M. 1984. Is there external digestion by Octopus? Journal qfZoology, beaks. London: Oxford University Press, 273pp. 202,441-447. Du,R.K 1959. Cephalopoda.British, AustralianandNewZealand Antarctic NIXON,M. 1987. Cephalopoddiets.InBone,P.R.,ed. CephalopodLifeCycles Research Expedition Reports, Series B, 8,89-106. Volume II. ComparativeReviews. London: Academic Press, 201-219. Dm,R.K. 197’2. Antarctic benthos. Advances in Marine Biology, 10,l-216. ROBSON,G.C. 1930. Cephalopoda. 1. Octopoda. Discovery R8eports,2, GILLER,P.S. 1984. Community structure and the niche. London: Chapman & 371-402. Hall, 176pp. ROBSON,G.C. 1932. A monograph of the recent Cephalopoda based on the GRISLEY,M.S. &Boa,P.R. 1990. Chitinase, anew enzyme inoctopussaliva. collections in the British Museum (Natural History). Part I1,The octopoda. Comparative Biochemistry and Physiology, 95B, 311-316. London: British Museum (Natural History). 359 pp. JOUSIN,M.L. 1905. Description de deux Eledones provenant de I’Exp6dition du ROPER,C.F.E. & Voss, G.L 1983. Guidelines for taxonomic descriptions of Dr Charcot dans 1’Antarctique. Memoires Societe Zoologique Francaise, cephalopod species. Memoirs of theNationa1 Museum Victoria, 414,49-63. 18,22-31. RYAN,B.F., Jom, B.L. &RYAN,T.A. 1985. Minitab Handbook, Boston: Krme, S. 1988. A contribution to the reproductive biology and geographical Duxbury Press, 379pp. distribution of Antarctic octopodidae (Cephalopoda). Malacologia, 29, TAKJ,I. 1961, On two new eledonid octopods from the Antarctic seas. Journal 89-100. of the Faculty of Fisheries andAnimal Husbandry, Hiroshima Universiy, MANGOLD,K 1987. Reproduction. inBoYLE,P.R., ed. Cephalopod life cycles, 3(2), 297-316. vol. II. London: Academic Press, 157-200. VOIGHT,J.R. 1991. Morphological variation in octopod specimens: reassessing MASSY,A.L. 1916. Mollusca Part 11. Cephalopods. British Antarctic “Terra the assumption of preservation-induced deformation. Malacologiu, 33, Nova” Expedition 1910: Zoology 2 (7), 151-161. 241-253. Nma, K. 1987. Cephalopods of the world. In LommAB. ed, Neptune City: T.F.W. Publishers Inc. Ltd., 351pp.